Urinary, Plasma, and Serum Biomarkers' Utility for Predicting Acute Kidney Injury Associated With Cardiac Surgery in Adults: A Meta-analysis

Abstract

Background: Early accurate detection of acute kidney injury (AKI) occurring after cardiac surgery may improve morbidity and mortality. Although several novel biomarkers have been developed for the early detection of AKI, their clinical utility in the critical intraoperative and immediate postoperative period remains unclear.

Study design: Systematic review and meta-analysis.

Setting & population: Adult patients having cardiac surgery.

Selection criteria for studies: EMBASE, CINAHL, Cochrane Library, Scopus, and PubMed from January 1990 until January 2015 were systematically searched for cohort studies reporting the utility of novel biomarkers for the early diagnosis of AKI after adult cardiac surgery. Reviewers extracted data for study design, population, timing of biomarker measurement and AKI occurrence, biomarker performance (area under the receiver operating characteristic curve [AUROC]), and risk of bias.

Index tests: Novel urine, plasma, and serum AKI biomarkers, measured intraoperatively and in the early postoperative period (<24 hours).

Reference tests: AKI was defined according to the RIFLE, AKIN, or 2012 KDIGO criteria.

Results: We found 28 studies reporting intraoperative and/or early postoperative measurement of urine (n=23 studies) or plasma or serum (n=12 studies) biomarkers. Only 4 of these studies measured biomarkers intraoperatively. Overall, intraoperative discrimination by the urine biomarkers neutrophil gelatinase-associated lipocalin (NGAL) and kidney injury marker 1 (KIM-1) demonstrated AUROCs<0.70, whereas N-acetyl-β-d-glucosaminidase (NAG) and cystatin C had AUROCs<0.75. In the immediate 24-hour postoperative period, the urine biomarkers NGAL (16 studies), KIM-1 (6 studies), and liver-type fatty acid binding protein (6 studies) exhibited composite AUROCs of 0.69 to 0.72. The composite AUROCs for postoperative urine cystatin C, NAG, and interleukin 18 were ≤0.70. Similarly, the composite AUROCs for postoperative plasma NGAL (6 studies) and cystatin-C (5 studies) were <0.70.

Limitations: Heterogeneous AKI definitions.

Conclusions: In adults, known urinary, plasma, and serum biomarkers of AKI possess modest discrimination at best when measured within 24 hours of cardiac surgery.

Keywords: Cardiopulmonary bypass; acute kidney injury; alpha-1-microglobulin (A1M); biomarker; diagnostic performance; early postoperative; hepcidin-25; intraoperative; liver-type fatty acid binding protein (L-FABP); monocyte chemoattractant protein 1 (CCL2); tumor necrosis factor alpha (TNF-α); uric acid.